Mutation Research, 232 (1990) 57-61 Elsevier
57
MUT 02110
The kinetics of D N A damage by bleomycin in mammalian cells Daniel L6pez-Larraza, Julio C. De Luca and N6stor O. Bianchi IMBICE, C.C. 403, 1900 La Plata (Argentina) (Accepted 25 February 1990)
Keywords: Bleomycin; DNA damage; Kinetics
Summary The kinetics of DNA damage by bleomycin (BLM) was assessed by measuring the amount of DNA breakage induced by BLM at different doses, treatment lengths, and treatment temperatures. DNA degradation was measured with the alkaline unwinding method. Comparison of the curves of DNA cleavage by BLM leads to the conclusion that low doses (1-5 ~tg/ml) and short treatments (5-15 min) produce marked damage in the DNA. High increases in BLM concentration produce relatively small increases in DNA damage above the levels obtained with low doses. Extension of treatment times does not increase the DNA degradation above the rate observed with 15-min treatments. The repair of DNA damage starts at about 15 min after the initiation of treatment. The mending of DNA breaks is very fast and extensive when BLM is no longer present. Repair not only implies the closing of DNA nicks, but very likely the degradation of the BLM molecules intercalated in the DNA interrupting the reactions responsible for the generation of free radicals. Persistence of BLM in the cell environment facilitates the replacement of degraded BLM molecules by new ones. This produces the persistent production of free radicals and the establishment of a balance between the processes of DNA damage and repair.
Bleomycin (BLM) is a glycopeptide antitumor antibiotic that is considered radiomimetic due to its capability to induce DNA damage by liberation of free radicals (Lown and Sim, 1977). In spite of this similarity, the mechanisms of action of ionizing radiations and BLM show substantial differences. Firstly, ionizing radiations do not bind to DNA, while BLM intercalates in the D N A and binds to divalent metal ions (Povirk et al., 1979; Miller et al., 1985). Secondly, radiation produces free radicals by direct degradation of water and
Correspondence: Dr. D. L6pez-Larraza, IMBICE, C.C. 403, 1900 La Plata (Argentina).
organic matter (Chapman, 1980); BLM, on the other hand, generates active oxygen species through mechanisms not yet well understood (Burger et al., 1981). Thirdly, DNA damage by radiation is random, while DNA sensitivity to BLM is influenced by the DNA base composition (Takeshita et al., 1978) and the structure of the chromatin fibril (Vig and Lewis, 1978). An additional difference between ionizing radiation and BLM is the velocity of damage to DNA. This event has been well studied for radiation (Chapman, 1980) and it is known to occur very fast ( - 1 × 10 -6 s). Conversely, although BLM damages DNA at a slower rate than radiations, there is scant information regarding the time se-
0027-5107/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
58
quence of the cellular events triggered by BLM. The aim of this report is to obtain further information on this matter. Material and methods
Chinese hamster ovary (CHO) cells were grown in Ham F10 medium supplemented with 10% fetal bovine serum and antibiotics. Aqueous stock solutions of BLM were prepared in such a way that addition of 0.1 ml to cell suspensions gave the required final concentrations. Discontinuation of BLM treatments was carried out by washing the cells twice with Hanks' saline or Tris-buffered solution (pH 7.4). Assessment of D N A cleavage was performed by the alkaline unwinding method (Rydberg, 1975). Exponentially growing CHO cells were labeled for 24 h with 0.1 /~Ci/ml of [3H]TdR (New England Nuclear, sp. act. 6.7 C i / m M ) and chased for 3 h with 2 x 10 -6 M cold TdR. Afterwards the cells were collected by trypsinization, washed and resuspended in fresh culture medium at a concentration of 10 6 cells/ml. BLM treatments were performed at this stage. At the end of treatments, the cells were washed twice with cold Tris-buffer solution. Aliquots of 10 6 cells/ml were lysed in 0.03 M NaOH, 0.01 M N a 2 H P O 4, 0.9 M NaC1, pH 12.2 for 30 min at room temperature in the dark. The samples were afterwards neutralized with an equal volume of 0.03 N HC1 and disrupted by sonication. Lysis was completed by adding 8.7 mM SDS. The samples were stored at - 2 0 ° C until further processing. Chromatography was performed in hydroxylapatite columns. Single-stranded (SS) and double-stranded (DS) DNAs were eluted with 0.125 M and 0.25 M potassium phosphate buffer respectively. The amounts of both forms of D N A in each experimental point were determined by liquid scintillation counting and the mass of D S D N A was obtained as follows: mass D S D N A (cpm) D S D N A (cpm) SSDNA + (cpm) D S D N A A decrease in the mass of D S D N A indicates D N A degradation (Rydberg, 1975).
Each experiment was performed in duplicate and the 2 sets of samples were run simultaneously. Moreover, 2-3 experiments (each one in duplicate) were carried out for each of the experimental models described in Results. The analysis of variance showed no significant differences ( p > 0.05) between the values obtained in each experimental point. Therefore, these values were averaged to construct the curves show in Figs. 1-3. Results
A total of 6 different dose-response curves were obtained. With no exception all curves showed the triphasic shape illustrated in Fig. 1. A drug concentration of 1-5 /~g/ml induced a marked D N A degradation. Between 5 and 10 /~g/ml, D N A degradation reached a plateau. At higher doses the D N A breakage started again, but at a much lower rate than with lower doses. The time-response curves of DNA degradation by BLM are depicted in Fig. 2. Once again all the curves analyzed (6 per experimental endpoint) showed the same pattern. A dose of 1 ffg/ml of BLM for 5 min at 37 o C failed to produce changes compared to control levels in the amount of D S D N A (Fig. 2A). On the other hand, 10 ffg/ml for the same time induced a noticeable quantity of
lol
0.9
0.8 < za
0.2 0.6
,..n 0.5 n 0.4 0.3 0.2
O.1 -oil
I
I
I
5 10
50
100
BLM (#g/rnl)
Fig. l. Changes in the mass of D S D N A induced by 1-100 ~ g / m l BLM. Note the marked slope of the curve at 1-5 ~ g / m l , the plateau at 5-10 ~ g / m l and the moderate decrease in D S D N A at high BLM doses (50-100 ~ g / m l ) .
59 C
A 0.8
0.7
0£ 0.5
01"i I 05
I
15
I 30
"rl I I 0 5 15
I 30
"~1 I
I
1 ..
0 5 15
30
Time (rain)
Fig. 2. Changes in the mass of D S D N A induced by different treatment times with 1 F g / m l (A) and 10 # g / m l (B) of BLM at 3 7 ° C and by 10 # g / m l at 4 ° C (C).
degraded DNA (Fig. 2B). Both doses produced maximal DNA damage in 15-rain treatments. Thereafter (30-min treatment at 37 o C), the rate of degradation showed a significant decrease (Fig. 2A,B). In treatments with 10 /~g/ml of BLM at 4 ° C the peak of DNA degradation was again observed at 15 min but no fall in the levels of DNA degradation at 30 min was detected (Fig. 2C). The kinetics of repair was analyzed as follows. At the end of a 15-min treatment with 10/~g/ml of BLM (37 o C) the drug was washed out (time 0, To) and the cells were reincubated at 37°C for 2.5, 5, 15 and 60 min. DNA cleavage was assessed at the end of each incubation period. The results obtained coincide with those from other authors (Hittelman et al., 1980) in showing that most of 1.0
0.9
0.8
< Z t/) c~
0.7
0.6 01" I I I TO 2.5 5 15 L._II BLM Repoir-time
I 60 I (min)
Fig. 3. Kinetics of D N A repair at 3 7 ° C (A) and 4 ° C (B) after a 15-min pulse treatment with BLM. Changes in the mass of DSDNA.
the DNA repair takes place during the initial 5 min (Fig. 3A). Thereafter, the closing of DNA nicks continues for 1 h or more at a slower rate than in the first phase (Fig. 3A). A clear modification in the repair-kinetics curve can be obtained by incubating the cells at 4 ° C after the BLM treatments (Fig. 3B). Under this condition DNA repair takes place for only 2.5 min, inducing extensive repair of the degraded DNA. At this moment (2.5 min) the repair system becomes inhibited by the low temperature and DNA cleavage starts again to reach values close to peak degradation values (To) at 15- and 60-min incubation times (Fig. 3B). Discussion
Data from mammalian cells indicate that BLM induces preferential DNA degradation in extended chromatin regions while condensed chromatin is affected to a lesser degree (Iqbal, 1976; Kuo and Hsu, 1979). CHO cells have extremely undercoiled chromatin regions and regions showing increasing levels of chromatin condensation. DNA from undercoiled chromatin is probably much exposed and can be extensively and rapidly degraded by low BLM doses (Fig. 1, 1-5 #g/ml). On the other hand, chromatin condensation very likely shields the DNA, making it necessary to increase the BLM concentration to high levels to produce a limited increment in DNA degradation (Fig. 1, 50-100 /xg/ml). Moderate increases in BLM dose are probably insufficient to damage the DNA in condensed chromatin and cannot induce further cleavage in undercoiled DNA. The result of this is the appearance of a plateau in the curve of DNA degradation (Fig. 1, 5-10 Fg/ml). At physiological temperatures the BLM damage to DNA is the final result of 4 different but interacting events: (a) DNA degradation by free radicals, (b) inactivation of free radicals by antioxidant enzymes, (c) enzymatic repair of damage, and (d) BLM degradation by BLM hydrolases (Umezawa et al., 1972). Low temperatures do not prevent event 'a' (Umezawa et al., 1973) but inhibit enzymatic events 'b-d'. Comparison of the curves in Fig. 2B, C shows that low temperature did not produce any substantial change in the initial response to 10 # g / m l
60 of BLM. This indicates that damage by free radicals is the most important (and perhaps the only) event determining the amount of D N A breakage during the initial 15 min of BLM action. Conversely, prolongation of treatments for 30 min at 3 7 ° C produced a paradoxical decrease of D N A degradation that was abolished by low temperatures (Fig. 2B, C). Therefore, it seems valid to conclude that the fall in D N A degradation occurring in long BLM treatments at physiological temperatures should be mediated by some of the enzymatic protective mechanisms enumerated above (b-d). Antioxidant enzymes are mainly located in the cytoplasm and are sufficient to inactivate the free radicals generated in that environment (Fridovich, 1978). On the other hand, BLM molecules intercalated in the D N A produce active oxygen species by reactions that occur in a microdomain of the chromatin, which affect only 1 D N A strand, and only the nucleotide located downstream from the point of BLM intercalation (Takeshita et al., 1978; Miller et al., 1985). Therefore, it seems improbable that antioxidant enzymes may produce any important control of free radicals generated by B L M / F e 2 + / D N A complexes. Accordingly, D N A repair and BLM hydrolases should be the 2 main mechanisms producing the decrease of D N A cleavage in long BLM treatments. The kinetics of D N A repair at 37 and 4 ° C was studied after a 15-min pulse treatment with BLM (the treatment was performed at 3 7 ° C and the drug was washed out after the treatment). When cells were placed at low temperatures the repair proceeded for only 2.5 min. Thereafter the damage to D N A started again reaching peak values in a short while (Fig. 3B). This result can only be interpreted by assuming a cooperative action of repair enzymes that close the D N A nicks and BLM hydrolases that degrade the drug. A progressive decrease in the cell temperature would decoupie the 2 enzymatic systems allowing the repair to continue for a few minutes but inhibiting the BLM degradation. Under this condition the drug remains intercalated in the D N A reinitiating the cleavage of the same sites that were repaired before D N A repair inhibition. Comparison of the curves of D N A repair in the presence and absence of BLM in the culture medium supports the above
assumption. Most of the D N A breakage by BLM is repaired within 5 min of discontinuation of the treatment (Fig. 3A). Conversely, the same period produces a slight repair of D N A nicks if the BLM remains in the culture medium (15-30-min component in the curve of Fig. 2B). These findings suggest that the persistence of BLM in the cell environment would allow new molecules to intercalate the D N A replacing the degraded ones. Such a mechanism would produce a continuous breakage of D N A that would be partially corrected by the repair system. Our results indicate that D N A damage by BLM is the result of several coincident processes that are triggered by the association of the drug with DNA. It is clear that low BLM doses with persistence of the drug in the cell environment induce marked D N A degradation. This conclusion may be relevant in planning the strategy for the clinical use of BLM: small drug doses at frequent intervals are expected to produce a better antitumor effect than high doses at longer intervals.
Acknowledgements This work was supported by grants from CONI C E T and CIC of Argentina. D. L6pez-Larraza was the recipient of a fellowship from CIC.
References Burger, R.M., J. Peisach and S.B. Horwitz (1981) Mechanism of bleomycin action: in vitro studies, Science, 208, 715-727. Chapman, D.J. (1980) Biophysical models of mammalian cell inactivation by radiation, in: R.E. Meyn and H.R. Withers (Eds.), Radiation Injury in Cancer Research, Raven Press, New York, pp. 21-32. Fridovich, I. (1978) The biology of oxygen radicals, Science, 201,875-880. Hittelman, W.N., M.A. Sognier and A. Cole (1980) Direct measurement of chromosome damage and its repair by premature chromosome condensation, in: R.E. Meyn and H.R. Withers (Eds.), Radiation Biology in Cancer Research, Raven Press, New York, pp. 103-123. Iqbal, Z.M., K.W. Kohn, R.A.G. Ewig and A.J. Fornace Jr. (1976) Single strand scission and repair of DNA in mammalian cells by bleomycin, Cancer Res., 36, 3834-3838. Kuo, M.T., and T.C. Hsu (1979) Biochemical and cytological studies of bleomycin actions on chromatin and chromosomes, Chromosoma, 68, 229-240. Lown, W.J., and S.K. Sim (1977) The mechanism of the bleomycin-induced cleavage of DNA, Biochem. Biophys. Res. Commun., 77, 1150-1157.
61 Miller, K.J., M. Lauer and W. Caloccia (1985) Interactions of molecules with nucleic acids. XII. Theoretical model for the interaction of a fragment of bleomycin with DNA, Biopolymers, 24, 913-934. Povirk, L.F., M. Hogan and N. Dattagupta (1979) Binding of bleomycin to DNA: intercalation of the bithiazole rings, Biochemistry, 19, 96-101. Rydberg, B. (1975) The rate of strand separation in alkali of DNA of irradiated mammalian cells, Radiat. Res., 61, 274-287. Takeshita, M., A.P. Grollman, E. Ohtsubo and H. Ohtsubo (1978) Interaction of bleomycin with DNA, Proc. Natl. Acad. Sci. (U.S.A.), 75, 5983-5987.
Umezawa, H., T. Takeuchi, S. Hori, T. Sawa and M. Ishizuka (1972) Studies on the mechanism of antitumor effect of bleomycin on squamous cell carcinoma, J. Antibiotics, 7, 409-420. Umezawa, H., H. Asakura, K. Oda and S. Hofi (1973) The effect of bleomycin on SV40 DNA: characteristics of bleomycin action which produces a single scission in a superhelical form of SV40 DNA, J. Antibiotics, 9, 521-527. Vig, B.K., and R. Lewis (1978) Genetic toxicology of bleomycin, Mutation Res., 55, 121-145.
57
MUT 02110
The kinetics of D N A damage by bleomycin in mammalian cells Daniel L6pez-Larraza, Julio C. De Luca and N6stor O. Bianchi IMBICE, C.C. 403, 1900 La Plata (Argentina) (Accepted 25 February 1990)
Keywords: Bleomycin; DNA damage; Kinetics
Summary The kinetics of DNA damage by bleomycin (BLM) was assessed by measuring the amount of DNA breakage induced by BLM at different doses, treatment lengths, and treatment temperatures. DNA degradation was measured with the alkaline unwinding method. Comparison of the curves of DNA cleavage by BLM leads to the conclusion that low doses (1-5 ~tg/ml) and short treatments (5-15 min) produce marked damage in the DNA. High increases in BLM concentration produce relatively small increases in DNA damage above the levels obtained with low doses. Extension of treatment times does not increase the DNA degradation above the rate observed with 15-min treatments. The repair of DNA damage starts at about 15 min after the initiation of treatment. The mending of DNA breaks is very fast and extensive when BLM is no longer present. Repair not only implies the closing of DNA nicks, but very likely the degradation of the BLM molecules intercalated in the DNA interrupting the reactions responsible for the generation of free radicals. Persistence of BLM in the cell environment facilitates the replacement of degraded BLM molecules by new ones. This produces the persistent production of free radicals and the establishment of a balance between the processes of DNA damage and repair.
Bleomycin (BLM) is a glycopeptide antitumor antibiotic that is considered radiomimetic due to its capability to induce DNA damage by liberation of free radicals (Lown and Sim, 1977). In spite of this similarity, the mechanisms of action of ionizing radiations and BLM show substantial differences. Firstly, ionizing radiations do not bind to DNA, while BLM intercalates in the D N A and binds to divalent metal ions (Povirk et al., 1979; Miller et al., 1985). Secondly, radiation produces free radicals by direct degradation of water and
Correspondence: Dr. D. L6pez-Larraza, IMBICE, C.C. 403, 1900 La Plata (Argentina).
organic matter (Chapman, 1980); BLM, on the other hand, generates active oxygen species through mechanisms not yet well understood (Burger et al., 1981). Thirdly, DNA damage by radiation is random, while DNA sensitivity to BLM is influenced by the DNA base composition (Takeshita et al., 1978) and the structure of the chromatin fibril (Vig and Lewis, 1978). An additional difference between ionizing radiation and BLM is the velocity of damage to DNA. This event has been well studied for radiation (Chapman, 1980) and it is known to occur very fast ( - 1 × 10 -6 s). Conversely, although BLM damages DNA at a slower rate than radiations, there is scant information regarding the time se-
0027-5107/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
58
quence of the cellular events triggered by BLM. The aim of this report is to obtain further information on this matter. Material and methods
Chinese hamster ovary (CHO) cells were grown in Ham F10 medium supplemented with 10% fetal bovine serum and antibiotics. Aqueous stock solutions of BLM were prepared in such a way that addition of 0.1 ml to cell suspensions gave the required final concentrations. Discontinuation of BLM treatments was carried out by washing the cells twice with Hanks' saline or Tris-buffered solution (pH 7.4). Assessment of D N A cleavage was performed by the alkaline unwinding method (Rydberg, 1975). Exponentially growing CHO cells were labeled for 24 h with 0.1 /~Ci/ml of [3H]TdR (New England Nuclear, sp. act. 6.7 C i / m M ) and chased for 3 h with 2 x 10 -6 M cold TdR. Afterwards the cells were collected by trypsinization, washed and resuspended in fresh culture medium at a concentration of 10 6 cells/ml. BLM treatments were performed at this stage. At the end of treatments, the cells were washed twice with cold Tris-buffer solution. Aliquots of 10 6 cells/ml were lysed in 0.03 M NaOH, 0.01 M N a 2 H P O 4, 0.9 M NaC1, pH 12.2 for 30 min at room temperature in the dark. The samples were afterwards neutralized with an equal volume of 0.03 N HC1 and disrupted by sonication. Lysis was completed by adding 8.7 mM SDS. The samples were stored at - 2 0 ° C until further processing. Chromatography was performed in hydroxylapatite columns. Single-stranded (SS) and double-stranded (DS) DNAs were eluted with 0.125 M and 0.25 M potassium phosphate buffer respectively. The amounts of both forms of D N A in each experimental point were determined by liquid scintillation counting and the mass of D S D N A was obtained as follows: mass D S D N A (cpm) D S D N A (cpm) SSDNA + (cpm) D S D N A A decrease in the mass of D S D N A indicates D N A degradation (Rydberg, 1975).
Each experiment was performed in duplicate and the 2 sets of samples were run simultaneously. Moreover, 2-3 experiments (each one in duplicate) were carried out for each of the experimental models described in Results. The analysis of variance showed no significant differences ( p > 0.05) between the values obtained in each experimental point. Therefore, these values were averaged to construct the curves show in Figs. 1-3. Results
A total of 6 different dose-response curves were obtained. With no exception all curves showed the triphasic shape illustrated in Fig. 1. A drug concentration of 1-5 /~g/ml induced a marked D N A degradation. Between 5 and 10 /~g/ml, D N A degradation reached a plateau. At higher doses the D N A breakage started again, but at a much lower rate than with lower doses. The time-response curves of DNA degradation by BLM are depicted in Fig. 2. Once again all the curves analyzed (6 per experimental endpoint) showed the same pattern. A dose of 1 ffg/ml of BLM for 5 min at 37 o C failed to produce changes compared to control levels in the amount of D S D N A (Fig. 2A). On the other hand, 10 ffg/ml for the same time induced a noticeable quantity of
lol
0.9
0.8 < za
0.2 0.6
,..n 0.5 n 0.4 0.3 0.2
O.1 -oil
I
I
I
5 10
50
100
BLM (#g/rnl)
Fig. l. Changes in the mass of D S D N A induced by 1-100 ~ g / m l BLM. Note the marked slope of the curve at 1-5 ~ g / m l , the plateau at 5-10 ~ g / m l and the moderate decrease in D S D N A at high BLM doses (50-100 ~ g / m l ) .
59 C
A 0.8
0.7
0£ 0.5
01"i I 05
I
15
I 30
"rl I I 0 5 15
I 30
"~1 I
I
1 ..
0 5 15
30
Time (rain)
Fig. 2. Changes in the mass of D S D N A induced by different treatment times with 1 F g / m l (A) and 10 # g / m l (B) of BLM at 3 7 ° C and by 10 # g / m l at 4 ° C (C).
degraded DNA (Fig. 2B). Both doses produced maximal DNA damage in 15-rain treatments. Thereafter (30-min treatment at 37 o C), the rate of degradation showed a significant decrease (Fig. 2A,B). In treatments with 10 /~g/ml of BLM at 4 ° C the peak of DNA degradation was again observed at 15 min but no fall in the levels of DNA degradation at 30 min was detected (Fig. 2C). The kinetics of repair was analyzed as follows. At the end of a 15-min treatment with 10/~g/ml of BLM (37 o C) the drug was washed out (time 0, To) and the cells were reincubated at 37°C for 2.5, 5, 15 and 60 min. DNA cleavage was assessed at the end of each incubation period. The results obtained coincide with those from other authors (Hittelman et al., 1980) in showing that most of 1.0
0.9
0.8
< Z t/) c~
0.7
0.6 01" I I I TO 2.5 5 15 L._II BLM Repoir-time
I 60 I (min)
Fig. 3. Kinetics of D N A repair at 3 7 ° C (A) and 4 ° C (B) after a 15-min pulse treatment with BLM. Changes in the mass of DSDNA.
the DNA repair takes place during the initial 5 min (Fig. 3A). Thereafter, the closing of DNA nicks continues for 1 h or more at a slower rate than in the first phase (Fig. 3A). A clear modification in the repair-kinetics curve can be obtained by incubating the cells at 4 ° C after the BLM treatments (Fig. 3B). Under this condition DNA repair takes place for only 2.5 min, inducing extensive repair of the degraded DNA. At this moment (2.5 min) the repair system becomes inhibited by the low temperature and DNA cleavage starts again to reach values close to peak degradation values (To) at 15- and 60-min incubation times (Fig. 3B). Discussion
Data from mammalian cells indicate that BLM induces preferential DNA degradation in extended chromatin regions while condensed chromatin is affected to a lesser degree (Iqbal, 1976; Kuo and Hsu, 1979). CHO cells have extremely undercoiled chromatin regions and regions showing increasing levels of chromatin condensation. DNA from undercoiled chromatin is probably much exposed and can be extensively and rapidly degraded by low BLM doses (Fig. 1, 1-5 #g/ml). On the other hand, chromatin condensation very likely shields the DNA, making it necessary to increase the BLM concentration to high levels to produce a limited increment in DNA degradation (Fig. 1, 50-100 /xg/ml). Moderate increases in BLM dose are probably insufficient to damage the DNA in condensed chromatin and cannot induce further cleavage in undercoiled DNA. The result of this is the appearance of a plateau in the curve of DNA degradation (Fig. 1, 5-10 Fg/ml). At physiological temperatures the BLM damage to DNA is the final result of 4 different but interacting events: (a) DNA degradation by free radicals, (b) inactivation of free radicals by antioxidant enzymes, (c) enzymatic repair of damage, and (d) BLM degradation by BLM hydrolases (Umezawa et al., 1972). Low temperatures do not prevent event 'a' (Umezawa et al., 1973) but inhibit enzymatic events 'b-d'. Comparison of the curves in Fig. 2B, C shows that low temperature did not produce any substantial change in the initial response to 10 # g / m l
60 of BLM. This indicates that damage by free radicals is the most important (and perhaps the only) event determining the amount of D N A breakage during the initial 15 min of BLM action. Conversely, prolongation of treatments for 30 min at 3 7 ° C produced a paradoxical decrease of D N A degradation that was abolished by low temperatures (Fig. 2B, C). Therefore, it seems valid to conclude that the fall in D N A degradation occurring in long BLM treatments at physiological temperatures should be mediated by some of the enzymatic protective mechanisms enumerated above (b-d). Antioxidant enzymes are mainly located in the cytoplasm and are sufficient to inactivate the free radicals generated in that environment (Fridovich, 1978). On the other hand, BLM molecules intercalated in the D N A produce active oxygen species by reactions that occur in a microdomain of the chromatin, which affect only 1 D N A strand, and only the nucleotide located downstream from the point of BLM intercalation (Takeshita et al., 1978; Miller et al., 1985). Therefore, it seems improbable that antioxidant enzymes may produce any important control of free radicals generated by B L M / F e 2 + / D N A complexes. Accordingly, D N A repair and BLM hydrolases should be the 2 main mechanisms producing the decrease of D N A cleavage in long BLM treatments. The kinetics of D N A repair at 37 and 4 ° C was studied after a 15-min pulse treatment with BLM (the treatment was performed at 3 7 ° C and the drug was washed out after the treatment). When cells were placed at low temperatures the repair proceeded for only 2.5 min. Thereafter the damage to D N A started again reaching peak values in a short while (Fig. 3B). This result can only be interpreted by assuming a cooperative action of repair enzymes that close the D N A nicks and BLM hydrolases that degrade the drug. A progressive decrease in the cell temperature would decoupie the 2 enzymatic systems allowing the repair to continue for a few minutes but inhibiting the BLM degradation. Under this condition the drug remains intercalated in the D N A reinitiating the cleavage of the same sites that were repaired before D N A repair inhibition. Comparison of the curves of D N A repair in the presence and absence of BLM in the culture medium supports the above
assumption. Most of the D N A breakage by BLM is repaired within 5 min of discontinuation of the treatment (Fig. 3A). Conversely, the same period produces a slight repair of D N A nicks if the BLM remains in the culture medium (15-30-min component in the curve of Fig. 2B). These findings suggest that the persistence of BLM in the cell environment would allow new molecules to intercalate the D N A replacing the degraded ones. Such a mechanism would produce a continuous breakage of D N A that would be partially corrected by the repair system. Our results indicate that D N A damage by BLM is the result of several coincident processes that are triggered by the association of the drug with DNA. It is clear that low BLM doses with persistence of the drug in the cell environment induce marked D N A degradation. This conclusion may be relevant in planning the strategy for the clinical use of BLM: small drug doses at frequent intervals are expected to produce a better antitumor effect than high doses at longer intervals.
Acknowledgements This work was supported by grants from CONI C E T and CIC of Argentina. D. L6pez-Larraza was the recipient of a fellowship from CIC.
References Burger, R.M., J. Peisach and S.B. Horwitz (1981) Mechanism of bleomycin action: in vitro studies, Science, 208, 715-727. Chapman, D.J. (1980) Biophysical models of mammalian cell inactivation by radiation, in: R.E. Meyn and H.R. Withers (Eds.), Radiation Injury in Cancer Research, Raven Press, New York, pp. 21-32. Fridovich, I. (1978) The biology of oxygen radicals, Science, 201,875-880. Hittelman, W.N., M.A. Sognier and A. Cole (1980) Direct measurement of chromosome damage and its repair by premature chromosome condensation, in: R.E. Meyn and H.R. Withers (Eds.), Radiation Biology in Cancer Research, Raven Press, New York, pp. 103-123. Iqbal, Z.M., K.W. Kohn, R.A.G. Ewig and A.J. Fornace Jr. (1976) Single strand scission and repair of DNA in mammalian cells by bleomycin, Cancer Res., 36, 3834-3838. Kuo, M.T., and T.C. Hsu (1979) Biochemical and cytological studies of bleomycin actions on chromatin and chromosomes, Chromosoma, 68, 229-240. Lown, W.J., and S.K. Sim (1977) The mechanism of the bleomycin-induced cleavage of DNA, Biochem. Biophys. Res. Commun., 77, 1150-1157.
61 Miller, K.J., M. Lauer and W. Caloccia (1985) Interactions of molecules with nucleic acids. XII. Theoretical model for the interaction of a fragment of bleomycin with DNA, Biopolymers, 24, 913-934. Povirk, L.F., M. Hogan and N. Dattagupta (1979) Binding of bleomycin to DNA: intercalation of the bithiazole rings, Biochemistry, 19, 96-101. Rydberg, B. (1975) The rate of strand separation in alkali of DNA of irradiated mammalian cells, Radiat. Res., 61, 274-287. Takeshita, M., A.P. Grollman, E. Ohtsubo and H. Ohtsubo (1978) Interaction of bleomycin with DNA, Proc. Natl. Acad. Sci. (U.S.A.), 75, 5983-5987.
Umezawa, H., T. Takeuchi, S. Hori, T. Sawa and M. Ishizuka (1972) Studies on the mechanism of antitumor effect of bleomycin on squamous cell carcinoma, J. Antibiotics, 7, 409-420. Umezawa, H., H. Asakura, K. Oda and S. Hofi (1973) The effect of bleomycin on SV40 DNA: characteristics of bleomycin action which produces a single scission in a superhelical form of SV40 DNA, J. Antibiotics, 9, 521-527. Vig, B.K., and R. Lewis (1978) Genetic toxicology of bleomycin, Mutation Res., 55, 121-145.
